AAAI 2024 论文列表

Generating realistic human motion sequences from text descriptions is a challenging task that requires capturing the rich expressiveness of both natural language and human motion. Recent advances in diffusion models have enabled significant progress in human motion synthesis. However, existing methods struggle to handle text inputs that describe complex or long motions. In this paper, we propose the Adaptable Motion Diffusion (AMD) model, which leverages a Large Language Model (LLM) to parse the input text into a sequence of concise and interpretable anatomical scripts that correspond to the target motion. This process exploits the LLM’s ability to provide anatomical guidance for complex motion synthesis. We then devise a two-branch fusion scheme that balances the influence of the input text and the anatomical scripts on the inverse diffusion process, which adaptively ensures the semantic fidelity and diversity of the synthesized motion. Our method can effectively handle texts with complex or long motion descriptions, where existing methods often fail. Experiments on datasets with relatively more complex motions, such as CLCD1 and CLCD2, demonstrate that our AMD significantly outperforms existing state-of-the-art models.

Authorship Attribution (AA) and Authorship Obfuscation (AO) are two competing tasks of increasing importance in privacy research. Modern AA leverages an author's consistent writing style to match a text to its author using an AA classifier. AO is the corresponding adversarial task, aiming to modify a text in such a way that its semantics are preserved, yet an AA model cannot correctly infer its authorship. To address privacy concerns raised by state-of-the-art (SOTA) AA methods, new AO methods have been proposed but remain largely impractical to use due to their prohibitively slow training and obfuscation speed, often taking hours. To this challenge, we propose a practical AO method, ALISON, that (1) dramatically reduces training/obfuscation time, demonstrating more than 10x faster obfuscation than SOTA AO methods, (2) achieves better obfuscation success through attacking three transformer-based AA methods on two benchmark datasets, typically performing 15% better than competing methods, (3) does not require direct signals from a target AA classifier during obfuscation, and (4) utilizes unique stylometric features, allowing sound model interpretation for explainable obfuscation. We also demonstrate that ALISON can effectively prevent four SOTA AA methods from accurately determining the authorship of ChatGPT-generated texts, all while minimally changing the original text semantics. To ensure the reproducibility of our findings, our code and data are available at: https://github.com/EricX003/ALISON.

Visual art facilitates expression, communication, and connection, yet it remains inaccessible to those who are visually-impaired and those who lack the resources to understand the techniques and history of art. In this work, I propose the development of a generative AI model that generates a description and interpretation of a given artwork. Such research can make art more accessible, support art education, and improve the ability of AI to understand and translate between creative media. Development will begin with a formative study to assess the needs and preferences of blind and low vision people and art experts. Following the formative study, the basic approach is to train the model on a database of artworks and their accompanying descriptions, predict sentiments from extracted visual data, and generate a paragraph closely resembling training textual data and incorporating sentiment analysis. The model will then be evaluated quantitatively through metrics like METEOR and qualitatively through Turing tests in an iterative process.

The application of artificial intelligence technology has greatly enhanced and fortified the safety of energy pipelines, particularly in safeguarding against external threats. The predominant methods involve the integration of intelligent sensors to detect external vibration, enabling the identification of event types and locations, thereby replacing manual detection methods. However, practical implementation has exposed a limitation in current methods - their constrained ability to accurately discern the spatial dimensions of external signals, which complicates the authentication of threat events. Our research endeavors to overcome the above issues by harnessing deep learning techniques to achieve a more fine-grained recognition and localization process. This refinement is crucial in effectively identifying genuine threats to pipelines, thus enhancing the safety of energy transportation. This paper proposes a radial threat estimation method for energy pipelines based on distributed optical fiber sensing technology. Specifically, we introduce a continuous multi-view and multi-domain feature fusion methodology to extract comprehensive signal features and construct a threat estimation and recognition network. The utilization of collected acoustic signal data is optimized, and the underlying principle is elucidated. Moreover, we incorporate the concept of transfer learning through a pre-trained model, enhancing both recognition accuracy and training efficiency. Empirical evidence gathered from real-world scenarios underscores the efficacy of our method, notably in its substantial reduction of false alarms and remarkable gains in recognition accuracy. More generally, our method exhibits versatility and can be extrapolated to a broader spectrum of recognition tasks and scenarios.

Effective teamwork translates to fewer preventable errors and higher task performance in collaborative tasks. However, in time-critical tasks, successful teamwork becomes highly challenging to attain. In such settings, often, team members have partial observability of their surroundings, incur high cost of communication, and have trouble estimating the state and intent of their teammates. To assist a team in improving teamwork at task time, my doctoral research proposes an automated task-time team intervention system. Grounded in the notion of shared mental models, the system first detects whether the team is on the same page or not. It then generates effective interventions to improve teamwork. Additionally, by leveraging past demonstrations to learn a model of team behavior, this system minimizes the need for domain experts to specify teamwork models and rules.

Generative artificial intelligence (AI) is swiftly cementing its role as an indispensable tool for students transitioning from K-12 to higher education and professional spheres. Yet, harnessing its full potential requires more than mere familiarity. Students must be equipped with the skills to engage with AI both productively and ethically. Left unchecked, AI usage can pose risks, especially if students lack proper guidance or understanding of their actions. Moreover, effective interaction with AI necessitates skills in prompt engineering to yield desired outcomes. Sidekick Academy is a digital online platform where students can safely experiment with and learn about AI. This article delves into the genesis of Sidekick Academy, offering a glimpse into its lessons on how to use AI and complex debate on ethical use. It also sheds light on the academy's "sandbox" - a secure space for students to explore AI without jeopardizing their safety or privacy.

As AI systems’ sophistication and proliferation have increased, awareness of the risks has grown proportionally. The AI industry is increasingly emphasizing the need for transparency, with proposals ranging from standardizing use of technical disclosures, like model cards, to regulatory licensing regimes. Since the AI value chain is complicated, with actors bringing varied expertise, perspectives, and values, it is crucial that consumers of transparency disclosures be able to understand the risks of the AI system in question. In this paper we propose a risk profiling standard which can guide downstream decision-making, including triaging further risk assessment, informing procurement and deployment, and directing regulatory frameworks. The standard is built on our proposed taxonomy of AI risks, which distills the wide variety of risks proposed in the literature into a high-level categorization. We outline the myriad data sources needed to construct informative Risk Profiles and propose a template and methodology for collating risk information into a standard, yet flexible, structure. We apply this methodology to a number of prominent AI systems using publicly available information. To conclude, we discuss design decisions for the profiles and future work.

Intelligent system audits are labor-intensive assurance activities that are typically performed once and discarded along with the opportunity to programmatically test all similar products for the market. This study illustrates how several incidents (i.e., harms) involving Named Entity Recognition (NER) can be prevented by scaling up a previously-performed audit of NER systems. The audit instrument's diagnostic capacity is maintained through a security model that protects the underlying data (i.e., addresses Goodhart's Law). An open-source evaluation infrastructure is released along with an example derived from a real-world audit that reports aggregated findings without exposing the underlying data.

In practical black-box attack scenarios, most of the existing transfer-based attacks employ pretrained models (e.g. ResNet50) as the substitute models. Unfortunately, these substitute models are not always appropriate for transfer-based attacks. Firstly, these models are usually trained on a largescale annotated dataset, which is extremely expensive and time-consuming to construct. Secondly, the primary goal of these models is to perform a specific task, such as image classification, which is not developed for adversarial attacks. To tackle the above issues, i.e., high cost and over-fitting on taskspecific models, we propose an Affordable and Generalizable Substitute (AGS) training framework tailored for transferbased adversarial attack. Specifically, we train the substitute model from scratch by our proposed adversary-centric constrastive learning. This proposed learning mechanism introduces another sample with slight adversarial perturbations as an additional positive view of the input image, and then encourages the adversarial view and two benign views to interact comprehensively with each other. To further boost the generalizability of the substitute model, we propose adversarial invariant learning to maintain the representations of the adversarial example invariants under augmentations with various strengths. Our AGS model can be trained solely with unlabeled and out-of domain data and avoid overfitting to any task-specific models, because of its inherently self-supervised nature. Extensive experiments demonstrate that our AGS achieves comparable or superior performance compared to substitute models pretrained on the complete ImageNet training set, when executing attacks across a diverse range of target models, including ViTs, robustly trained models, object detection and segmentation models. Our source codes are available at https://github.com/lwmming/AGS.

Audio-driven talking head synthesis is a promising topic with wide applications in digital human, film making and virtual reality. Recent NeRF-based approaches have shown superiority in quality and fidelity compared to previous studies. However, when it comes to few-shot talking head generation, a practical scenario where only few seconds of talking video is available for one identity, two limitations emerge: 1) they either have no base model, which serves as a facial prior for fast convergence, or ignore the importance of audio when building the prior; 2) most of them overlook the degree of correlation between different face regions and audio, e.g., mouth is audio related, while ear is audio independent. In this paper, we present Audio Enhanced Neural Radiance Field (AE-NeRF) to tackle the above issues, which can generate realistic portraits of a new speaker with few-shot dataset. Specifically, we introduce an Audio Aware Aggregation module into the feature fusion stage of the reference scheme, where the weight is determined by the similarity of audio between reference and target image. Then, an Audio-Aligned Face Generation strategy is proposed to model the audio related and audio independent regions respectively, with a dual-NeRF framework. Extensive experiments have shown AE-NeRF surpasses the state-of-the-art on image fidelity, audio-lip synchronization, and generalization ability, even in limited training set or training iterations.

Graph anomaly detection is crucial for identifying nodes that deviate from regular behavior within graphs, benefiting various domains such as fraud detection and social network. Although existing reconstruction-based methods have achieved considerable success, they may face the Anomaly Overfitting and Homophily Trap problems caused by the abnormal patterns in the graph, breaking the assumption that normal nodes are often better reconstructed than abnormal ones. Our observations indicate that models trained on graphs with fewer anomalies exhibit higher detection performance. Based on this insight, we introduce a novel two-stage framework called Anomaly-Denoised Autoencoders for Graph Anomaly Detection (ADA-GAD). In the first stage, we design a learning-free anomaly-denoised augmentation method to generate graphs with reduced anomaly levels. We pretrain graph autoencoders on these augmented graphs at multiple levels, which enables the graph autoencoders to capture normal patterns. In the next stage, the decoders are retrained for detection on the original graph, benefiting from the multi-level representations learned in the previous stage. Meanwhile, we propose the node anomaly distribution regularization to further alleviate Anomaly Overfitting. We validate the effectiveness of our approach through extensive experiments on both synthetic and real-world datasets.

Decision Transformer (DT), which employs expressive sequence modeling techniques to perform action generation, has emerged as a promising approach to offline policy optimization. However, DT generates actions conditioned on a desired future return, which is known to bear some weaknesses such as the susceptibility to environmental stochasticity. To overcome DT's weaknesses, we propose to empower DT with dynamic programming. Our method comprises three steps. First, we employ in-sample value iteration to obtain approximated value functions, which involves dynamic programming over the MDP structure. Second, we evaluate action quality in context with estimated advantages. We introduce two types of advantage estimators, IAE and GAE, which are suitable for different tasks. Third, we train an Advantage-Conditioned Transformer (ACT) to generate actions conditioned on the estimated advantages. Finally, during testing, ACT generates actions conditioned on a desired advantage. Our evaluation results validate that, by leveraging the power of dynamic programming, ACT demonstrates effective trajectory stitching and robust action generation in spite of the environmental stochasticity, outperforming baseline methods across various benchmarks. Additionally, we conduct an in-depth analysis of ACT's various design choices through ablation studies. Our code is available at https://github.com/LAMDA-RL/ACT.

Data augmentation plays a crucial role in improving the data efficiency of reinforcement learning (RL). However, the generation of high-quality augmented data remains a significant challenge. To overcome this, we introduce ACAMDA (Adversarial Causal Modeling for Data Augmentation), a novel framework that integrates two causality-based tasks: causal structure recovery and counterfactual estimation. The unique aspect of ACAMDA lies in its ability to recover temporal causal relationships from limited non-expert datasets. The identification of the sequential cause-and-effect allows the creation of realistic yet unobserved scenarios. We utilize this characteristic to generate guided counterfactual datasets, which, in turn, substantially reduces the need for extensive data collection. By simulating various state-action pairs under hypothetical actions, ACAMDA enriches the training dataset for diverse and heterogeneous conditions. Our experimental evaluation shows that ACAMDA outperforms existing methods, particularly when applied to novel and unseen domains.

The Aesthetics Assessment of Children's Paintings (AACP) is an important branch of the image aesthetics assessment (IAA), playing a significant role in children's education. This task presents unique challenges, such as limited available data and the requirement for evaluation metrics from multiple perspectives. However, previous approaches have relied on training large datasets and subsequently providing an aesthetics score to the image, which is not applicable to AACP. To solve this problem, we construct an aesthetics assessment dataset of children's paintings and a model based on self-supervised learning. 1) We build a novel dataset composed of two parts: the first part contains more than 20k unlabeled images of children's paintings; the second part contains 1.2k images of children's paintings, and each image contains eight attributes labeled by multiple design experts. 2) We design a pipeline that includes a feature extraction module, perception modules and a disentangled evaluation module. 3) We conduct both qualitative and quantitative experiments to compare our model's performance with five other methods using the AACP dataset. Our experiments reveal that our method can accurately capture aesthetic features and achieve state-of-the-art performance.

We describe our methodology for classifying ASL (American Sign Language) gestures. Rather than operate directly on raw images of hand gestures, we extract coor-dinates and render wireframes from individual images to construct a curated training dataset. This dataset is then used in a classifier that is memory efficient and provides effective performance (94% accuracy). Because we con-struct wireframes that contain information about several angles in the joints that comprise hands, our methodolo-gy is amenable to training those interested in learning ASL by identifying targeted errors in their hand gestures.

Currently, students acquire driving skills by practicing in actual traffic conditions and through direct interactions with an instructor. While one-on-one interactions could be tailored to a student’s learning style and skill level, making them effective for learning, one-on-one interactions are also inefficient, potentially costly, and not standardized with limitations on which traffic situation can be safely taught. For these exact reasons Way AS has developed and commercially deployed a virtual driving instructor that educates students in high-fidelity simulators. In this paper, we present a module, the Lesson generator, that extends the virtual driving instructor to generate personalized lessons for individual students with the goal to practice in a focused and deliberately fashion the skills that need practice for the students to become proficient drivers. A case study is presented, and the path to deployment is discussed.

Continuously observed event occurrences, often exhibit self and mutually exciting effects, which can be well modeled using temporal point processes. Beyond that, these event dynamics may also change over time, with certain periodic trends. We propose a novel variational autoencoder to capture such a mixture of temporal dynamics. More specifically, the whole time interval of the input sequence is partitioned into a set of sub intervals. The event dynamics are assumed to be stationary within each subinterval, but could be changing across those subintervals. In particular, we use a sequential latent variable model to learn a dependency graph between the observed dimensions, for each subinterval. The model predicts the future event times, by using the learned dependency graph to remove the non contributing influences of past events. By doing so, the proposed model demonstrates its higher accuracy in predicting inter event times and event types for several real world event sequences, compared with existing state of the art neural point processes.

Well-designed prompts have demonstrated the potential to guide text-to-image models in generating amazing images. Although existing prompt engineering methods can provide high-level guidance, it is challenging for novice users to achieve the desired results by manually entering prompts due to a discrepancy between novice-user-input prompts and the model-preferred prompts. To bridge the distribution gap between user input behavior and model training datasets, we first construct a novel Coarse-Fine Granularity Prompts dataset (CFP) and propose a novel User-Friendly Fine-Grained Text Generation framework (UF-FGTG) for automated prompt optimization. For CFP, we construct a novel dataset for text-to-image tasks that combines coarse and fine-grained prompts to facilitate the development of automated prompt generation methods. For UF-FGTG, we propose a novel framework that automatically translates user-input prompts into model-preferred prompts. Specifically, we propose a prompt refiner that continually rewrites prompts to empower users to select results that align with their unique needs. Meanwhile, we integrate image-related loss functions from the text-to-image model into the training process of text generation to generate model-preferred prompts. Additionally, we propose an adaptive feature extraction module to ensure diversity in the generated results. Experiments demonstrate that our approach is capable of generating more visually appealing and diverse images than previous state-of-the-art methods, achieving an average improvement of 5% across six quality and aesthetic metrics. Data and code are available at https://github.com/Naylenv/UF-FGTG.

Answer Set Programming (ASP) is a prominent rule-based language for knowledge representation and reasoning with roots in logic programming and non-monotonic reasoning. The aim to capture the essence of removing (ir)relevant details in ASP programs led to the investigation of different notions, from strong persistence (SP) forgetting, to faithful abstractions, and, recently, strong simplifications, where the latter two can be seen as relaxed and strengthened notions of forgetting, respectively. Although it was observed that these notions are related, especially given that they have characterizations through the semantics for strong equivalence, it remained unclear whether they can be brought together. In this work, we bridge this gap by introducing a novel relativized equivalence notion, which is a relaxation of the recent simplification notion, that is able to capture all related notions from the literature. We provide the necessary and sufficient conditions for relativized simplifiability, which shows that the challenging part is for when the context programs do not contain all the atoms to remove. We then introduce an operator that combines projection and a relaxation of SP-forgetting to obtain the relativized simplifications. We furthermore provide complexity results that complete the overall picture.